The Canada Basin 1989-1995 : upstream events and far-field effects of the Barents Sea Branch

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Bell & Howell Information and Learning

300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600

by

Fiona Ann McLaughlin B.Sc. University of Victoria, 1972 M.Sc., University of Victoria, 1996

A Dissertation Submitted in Partial Fulfilment of the Requirements for the Degree of

DOCTOR O F PHILOSOPHY

in the School of Earth and Ocean Sciences W e accept this dissertation as conforming

to the required standard

Dr. A. J. Weaver, Supervisor (School of Earth and Ocean Sciences)

Dr. E. C. Carmack, Departmental Member (School of Earth and Ocean

Dr. R. W . Macdonald, Departmental Member (School of Earth and Ocean Sciences)

Dr. N.n'urner, Outside Member (Department of Environmental Sciences) _________________

Dr. K. K. Falkner, External Examiner (Oregon State University)

© Fiona Ann McLaughlin, 2000 University of Victoria

All rights reserved. This thesis may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.

Abstract

Physical and geochemical tracer measurements were collected at one oceanographic station (Station A: 72 N 143 W) in the southern Canada Basin from 1989 to 1995, along sections from the Beaufort Shelf to this station in 1993 and 1995, and along a section westward of Banks Island in 1995. These measurements were examined to see how recent events in three upstream Arctic Ocean sub-basins impacted upon Canada Basin waters. Upstream events included Atlantic layer warming, relocation of the Atlantic/Pacific water mass boundary, and increased ventilation of boundary current waters. Early signals of change appeared first in the Canada Basin in 1993 along the continental margin and, by 1995, were evident at Station A in the basin interior and farther downstream. Differences in physical and geochemical properties (nutrients, oxygen, ’^1 and CFCs) were observed throughout much of the water column to depths greater than 1600 m. In particular, the boundary distinguishing Pacific from Atlantic-origin water was found to be shallower and Atlantic-origin water occupied more of the Canada Basin water column. By 1995, Atlantic-origin water in the lower halocline at Station A was found to be colder and more ventilated. Likewise, within the Atlantic layer, Fram Strait Branch (FSB) water was colder, fresher, and more ventilated, and Barents Sea Branch (BSB) water was warmer, fresher, and more ventilated than during previous years. By comparing observations at Station A with eastern Nansen

Basin observations, the main source of these changes was traced to dense water outflow from the Barents Sea. Studies indicated that in early 1989 Barents Sea waters were 2 °C warmer and that, between 1988 and 1989, a large volume of dense water had left the shelf. These events coincided with an atmospheric shift to increased cyclonic circulation in 1989, a transition unprecedented in its magnitude, geographic reach, and apparent oceanographic impact. The effects of a large outflow of dense Barents Sea water were observed some 5000 km away downstream in the Canada Basin where the BSB component of the Atlantic layer had increased 20% by 1995. Examirerfc:

Dr_A. Ji w eaver. Supervisor (School of Earth and Ocean Sciences)

Dr. E. C.^Carmack, Departmental Member (School of Earth and Ocean Scieno

Dr. R. W. Macdonald, Departmental Member (School of Earth and Ocean Sciences)

Dr. N. Turnèh Outside Member (Department of Environmental Sciences) _______________________________ Dr. K. K. Falkner, External Examiner (Oregon State University)

Table of Contents

Abstract... ii

Table of Contents... iv

Acknowledgements

...viii

1. Introduction... 1

2. Methods...5

3. Observations...7

3.1 W ater column description...7

3.2 Temperature and salinity observations...10

3.3 Barents Sea outflow characteristics...18

3.4 Time-series geochemical observations... 22

3.5 Canada Basin sections... 40

4. Discussion...53

4.1 Displacement of Pacific-origin w ater... 53

4.2 Change in Atlantic-origin water... 58

4.3 Upstream events...61

4.3.1 Barents Sea atmospheric forcing... 62

4.3.2 Variability and dense water formation in the Barents S e a ... 64

4.4 Transport of upstream change...6 7 4.4.1 Boundary current ra te ... 69

4.5 Arctic atmosphere-ocean system... 73

5. Conclusion...79

List of Figures

Figure 1. Arctic Ocean map Illustrating bathymetric features and circulation of Inflowing Atlantic (red) and Pacific (blue) waters. Map inset shows the

Canada Basin study area...2

Figure 2a. CTD potential temperature versus salinity from Canada Basin Station A, 1989-1995: 8=25 to 8 = 3 5 ... 11

Figure 2b. CTD potential temperature versus salinity: 8=31.0 to 8 = 3 4 .7 ... 12

Figure 2c. CTD potential temperature versus salinity: 8=34.75 to 34.96...13

Figure 3. CTD potential temperature profile at Canada Basin Station A, 1989-1995... 15

Figure 4. CTD Salinity profile at Canada Basin Station A, 1989-1995...16

Figure 5. CTD potential temperature versus salinity: 8=34.78 to 8=34.98. Data are from Canada Basin Station A (1989-1995); Nansen Basin Stations 21 and 70, (1993); Amundsen Basin Station 35 (1993); and Makarov Basin Stations E (1993) and 20 (1994)... 19

Figure 6. CTD potential temperature versus salinity: S=33.8 to 8=35.0. Data are from Canada Basin Station A (1989-1995); Nansen Basin Stations 21 and 70, (1993); Amundsen Basin Station 35 (1993); and Makarov Basin Stations E (1993) and 20 (1994)... 21

Figure 7. Silicate, ’^1 and saliniity profiles at Canada Basin Station A, 1995.23 Figure 8. Silicate profiles (a) 0 m to 500 m (b) 500 m to 2000 m... 24

Figure 9. Silicate versus salinity (a) 8=31 to S=35 (b) 8 = 34.5 to 8 = 3 5 .0 ... 26

Figure 10. Nitrate profiles... 28

Figure 11. Nitrate versus salinity... 29

Figure 12. Phosphate profiles...30

Figure 13. Phosphate versus salinity... 31

Figure 14. Oxygen profiles... 33

Figure 15. Oxygen versus salinity...34

Figure 16. CFC-11 profiles... 35

Figure 17. C F C -1 1 versus salinity... 36

Figure 18. '^ I profiles at Canada and Makarov basins...38

Figure 20a. CTD potential temperature versus salinity, 8=33.0 to 8 = 3 4 .4 .Data are from Canada Basin section stations in 1993 and 1995... 41

Figure 20b. CTD potential temperature versus salinity, 8 =34.80 to 8=34.92.. 42

Figure 21. C F C -1 1 profiles from 1993 and 1995 Canada Basin section stations. ...45

Figure 22. C F C -1 1 versus salinity from 1993 and 1995 Canada Basin section stations...46

Figure 23. '^1 profiles from 1995 Canada Basin section stations...49

Figure 24. CTD potential temperature profiles from 1995 Canada Basin section stations... 51

Figure 26. Silicate versus salinity correlation from 1992 Canada Basin Station A and Oden-91 stations located near the Morris Jesup Plateau...57 Figure 27. Cross-sectional representation of the boundary current as observed

along the southern Canada Basin... 68 Figure 29. C F C -1 1 age profiles from 1995 Canada Basin section stations 72 Figure 30. Two Arctic atmosphere-ocean modes; a hypothesis... 75 Figure 31. Relationship between Arctic atmosphere circulation, Arctic Ocean

Acknowledgements

I am fortunate to have had the opportunity to learn about oceanography in the field and in the classroom. I would like to thank Andrew W eaver for supervising my doctoral program, for his thoughtful guidance, and for treating me always as a colleague. Special thanks are due to Eddy Carmack and Robie Macdonald at the Institute of Ocean Sciences who encouraged me to pursue my interests into the “why” of things and to return to university. I am particularly indebted to Eddy for teaching me that physical oceanography could be grasped without having a fondness for equations, to think in temperature-salinity space, and to see ocean processes in “pictures”. I am no less indebted to Robie for teaching me by example the dedication that the study and practice of oceanography requires and for his help in demystifying geochemical models.

I have also been instructed in oceanography by the captains and crews of the Canadian Coast Guard icebreakers, especially Captains Gomes and Grandy, and the men and women who sailed with them. I have also learned much from my colleagues at the Institute of Ocean Sciences whose work in sampling and analysis contributed enormously to the data sets upon which this study of the Canada Basin is based. John Smith at the Bedford Institute was particularly helpful and supplied the iodine isotope data used here.

Last but not least, I would like to thank my partner, Thomas Fleming, who has taught me that the questions of research are universal, regardless of discipline, and that all researchers must first understand the “story” of their investigation in order to tell it to others, directly and simply. Together, all these people have taught me, as the novelist Laurence Sterne wrote in Tristram Shandv. ” There is a North-west passage to the intellectual world.”

Over the past decade, the historical view of the Arctic Ocean as an ocean in a steady state (c.f. Coachman and Barnes, 1961; Treshnikov, 1959; Carmack, 1986) has given way to a newer view of the Arctic as an ocean in a state of transition. This new understanding is a product of several factors. Science-capable icebreakers and submarines provided new platforms for discovery and greatly expanded the geography of Arctic inquiry from basin-wide sampling in the 1980s to ocean-wide expeditions in the 1990s. Modem geochemical measurements also furnished more complete descriptions about the composition of arctic waters and how these waters are interconnected. Since the late 1970s, atmospheric, biological, and modelling studies have also expanded the scope of scientific research, making it possible to test the validity of old understandings and to consider new questions. A growing literature now sustains comparison among high-quality data sets assembled over time and over a larger domain. This enriched knowledge base has provided a portrait of the Arctic as a dynamic ocean, and one that is an integral part of the larger global ocean system.

Views of a changing Arctic Ocean were first outlined in several studies documenting temperature change in the water column of three of the ocean’s four main sub-basins (see Figure 1). Warmer temperatures in the Atlantic layer of the water column were initially obsen/ed in 1990 in the Nansen Basin, geographically the first Arctic Ocean basin to receive Atlantic water inflow (Quadfasel et al., 1991; Quadfasel et al., 1993). Downstream of this inflow, Atlantic layer temperatures warmer than the climatological record were obsen/ed in 1993 over the Lomonosov Ridge in the Amundsen Basin (Morison et al., 1998) and, also, near the Mendeleyev Ridge in the southern Makarov Basin (Carmack et al., 1995). Warmer Atlantic layer temperatures were further confirmed by comparison of 1991 and 1994 temperatures on the eastern flank of the Lomonosov Ridge, which showed an increase of 0.5-0.6°C over this brief period (Swift et al., 1998).

B e a u f o r t S e a

Siberia

f

/

'^

t

Greenland

Alaska yBeaufoi\

Canada

Figure 1. Arctic Ocean map illustrating bathymetric features and circulation of Pacific (blue) waters. Map inset shows the Canada Basin study area.

m s »

inflowing Atlantic (red) and

circulation of the Arctic Ocean itself. The historical view of the Arctic Ocean (Gorshkov, 1983; Moore et al., 1983) held that the Eurasian Basin halocline consisted of Atlantic-origin water characterized by cold temperatures and low nutrients. Further, it held that the Canadian Basin halocline was comprised largely of Pacific-origin water, characterized by a local temperature maximum and minimum as well as a nutrient maximum (Kinney et al., 1970). Observations showed that these two haloclines were separated by a boundary located above the Lomonosov Ridge. This view was first questioned in 1991 when silicate concentrations in the upper layer of the Makarov Basin near the Lomonosov Ridge were found to be markedly lower than previously reported. This finding suggested the halocline front was variable (Anderson et al., 1994). The historical view was again challenged when low-nutrients were observed in the southern Makarov Basin halocline in 1993, a finding that signalled an eastward shift of the Atlantic/Pacific water mass boundary from a location over the Lomonosov Ridge to one over the Mendeleyev Ridge (McLaughlin et al., 1996). In addition to displacement of Pacific- origin water in the halocline, and the concomitant shallowing of the Atlantic layer, southem Makarov Basin waters were also found to be fresher and more ventilated between 400 m and 1600 m than in the overlying Atlantic layer. Increased ventilation suggested that rapid transport of upstream shelf-water by topographically-steered boundary currents had occurred. Observations of Atlantic- origin water extending from the Mendeleyev Ridge to the Lomonosov Ridge across the Makarov Basin in 1994 confirmed the basin-wide extent of the Atlantic/Pacific water mass transition (Carmack et al., 1998). Presence of ventilated water at depths greater than 1500 m in basin slope locations likewise confirmed transport of water from shelf regions (Swift et al., 1998).

Arctic Ocean research has focused principally on investigations of physical and geochemical changes in the Nansen, Amundsen and Makarov basins. Investigation into these three basins (hereafter referred to as “upstream” because of their pre-eminent location with respect to Atlantic water inflow, illustrated in

downstream In the Canada Basin— the fourth of the arctic basins, the largest, and the one which most directly influences outflow to the Canadian Archipelago and Labrador Sea. TIme-series study of the Canada Basin to date has been limited to two Investigations. Initial work was based on temperature and salinity measurements collected in the southem Canada Basin between 1979 and 1996 (Mailing, 1998). Mailing reported that changes were progressive during these 17 years. Compared to the early 1980s, the upper halocline was warmer, the lower halocline was colder, the Atlantic warm core was colder and fresher and, below this core, waters were warmer and fresher. He attributed the difference in the upper halocline to a decrease in strength of the Siberian High, and the difference in the lower halocline and increased thickness of the Atlantic layer's warm core to variation in flow via the Barents Sea.

The second investigation into the Canada Basin presented measurements collected in the Beaufort Sea between 1987 and 1997 (Macdonald et al., 1999). Resolving freshwater in the water column’s top 250 m and comparing sea-ice melt and runoff components, this study reported that sea-ice melt increased by 4-6 m after 1989. Macdonald et al. (1999) noted that the timing of this increase coincided with three events— a shift in the Arctic Oscillation Index (Overland et al., 1999), reduced anti-cyclonic wind forcing (Proshutinsky and Johnson, 1997), and a record minimum in ice extent (Serreze et al., 1995). Because increase in sea-ice melt was unaccompanied by change in runoff, they proposed that sea-ice melt increase was a thermal and mechanical response to the 1989 atmospheric- circulation regime shift.

Increased atmospheric cyclonicity and the resulting adjustment of ice distribution and thickness also serves to explain obsen/ations of anomalously thin multi-year ice in the Beaufort Sea in 1997 by McPhee et al. (1998) and Welch (1998). Comparison of sea-ice draft measurements from the larger Arctic Ocean between 1958-1976 and 1993-1997 substantiated that sea-ice thickness in the Beaufort and Chukchi seas decreased, and also identified that the decrease was greatest in the central and eastem Arctic (Rothrock et al., 1999). A reduction in the

Vinnikkov et al. (1999) to conclude that obsen/ed decreases in sea-ice extent were far greater than could be expected from natural climate variations and were related to anthropogenic global warming.

The following paper attempts to record differences in Canada Basin waters from 1989 to 1995 and to connect such developments to the broader Arctic Ocean which is, in turn, part of a complex global oceanic and atmospheric system. It will do so, first, by examining physical and geochemical measurements for evidence of change at one deep basin station in the southem Canada Basin. It will then compare these measurements to physical and geochemical measurements collected in upstream basins during three expeditions—the 1993 southem Makarov Basin expedition, the 1993 southeastern Eurasian Basin ARK IX/4 expedition, and the 1994 Arctic Ocean section expedition. Finally, measurements from sections extending from the Canada Basin shelf to the basin interior will be compared by time and location. Variation in water mass composition will be determined by geochemical tracers. For example: nutrients and ’” l will be used to define the intersection of Pacific and Atlantic waters within the water column; and, oxygen and CFC-11 will be used to distinguish between shelf and non-shelf sources of Atlantic layer water.

In summary, the following discussion will attempt to answer three questions: What changes have taken place in the Canada Basin since 1989? Can they be attributed to a source? And, how can they be explained?

2. Methods

station A was located at 72 ° N 143 ° W in the southem Canada Basin at a water column depth of 3300 m. Oceanographic samples were first collected in the summer of 1989 and, subsequently, collected every autumn with the exception of 1991 and 1994. In addition, oceanographic samples were collected along two sections in the Beaufort Sea, one northward from the Tuktoyatuk Peninsula to

interior to a depth of 1200 m in 1995 (Figure 1).

CTD data were collected from 1989 to 1992 with a Guildline conductivity- temperature-depth (CTD) profiler and, in 1993 and 1995, using a Falmouth Scientific Instruments ICTD. Measurements of temperature, pressure, and conductivity were recorded during downcasts. Instruments were calibrated before and after each cruise. Potential temperatures (6) and densities (a) were calculated using UNESCO (1963) algorithms.

Water samples were collected between 1989 and 1992 with 5 and 10 L Niskin bottles deployed from a wire and, in 1993 and 1995, collected in 10 L Niskin- type bottles mounted on a General Oceanics rosette together with the FSI ICTD and tripped on the upcast. Subsamples drawn for salinity (S), oxygen and nutrients were analyzed onboard. Halocarbon samples, collected in 1992, 1993, and 1995, were also analyzed onboard. Samples for '^1 measurements were collected in 1993 and 1995 for later analysis. A complete discussion of methods was reported in the Canadian Data Report Hydrography and Ocean Sciences data report series 1990-1996. Each year, the same analytical methods were used and the precision reported below, based on 1995 measurements, typified the entire time period.

Salinities were determined from conductivity measurements using a Guildline Autosal salinometer and referenced to Standard Sea Water, lAPSO Batch P118 (pooled variance Sp=0.0012, n=33). Oxygen was analyzed by the Carpenter-Winkler method (Sp=1.57 mmol m"^, n=85); silicate and nitrate by standard AutoAnalyzer Technicon methods, and phosphate by a modified method (Si(0 H)4: Sp=0.16 mmol m^, n=496; NO3: Sp=0.15 mmol m*^, n=493; PO4: Sp=0.06 mmol m"^, n=502). Halocarbons were analyzed by an automated purge and trap system combined with a Hewlett-Packard gas chromatograph electron capture detector (CFC-12: Sp=0.07 nmol m^, n=26; CFC-11: Sp=0.07 nmol m"^, n=26; CFC-113: Sp=0.06 nmol m"^, n=27; and CCU: Sp=0.24 nmol m"^, n=26). Only CFC-11 and CFC-12 were analyzed in 1992 whereas, in 1993 and in 1994, all four halocarbons were analyzed. All concentrations were reported using the SI093

Smith et al., 1998) at the IsoTrace Laboratory at the University of Toronto. Sample data were normalized to IsoTrace Reference Material #2 ('^1/ '^^1 = 1.174x10 ” atom ratio). The blank for this procedure is 0.75 ± 0.10x10® atoms/liter and the standard deviation (one sigma) ranged from 5-10% (Edmonds et al., 1998).

To remove inter-year offsets, all deep casts (below 1700 m) were internally calibrated. This was undertaken with the assumption that modification occurs extremely slowly in the 450-500 year-old deep layer (Macdonald et al., 1993; Schlosser et al., 1997) and would be undetectable over periods as short as six years. Accordingly, the deep theta minimum (2500 m) and maximum (2950 m) were assumed to occur at the same temperature and salinity given the comparatively short time-scale of this study. The magnitude of correction applied was S=<0.001 and 0 =<0.005 °C, except for 1992 when the correction was 0.01 °C. Geochemical data were internally calibrated using the same approach. The magnitude of these corrections, which were similar in size to the standard error for each measurement, was N0 3=<0 .4 mmol m"®, P04=<0.04 mmol m"®, Si(OH)4=<0.3 mmol m"®, and 02=<1 mmol m*®. These corrections were approximately an order of magnitude smaller than differences identified in the discussion below (i.e. A6=0.03 “C, aS=0.01, AN0 3 = 2 mmol m"®, APO4 =0.1 mmol m"®, ASi(OH)4=2-4 mmol m"®, AOa=20 mmol m"®).

3. Observations

3.1 Water column description

The Arctic Ocean water column consists of three main layers: a cold, low-salinity upper layer; a warm Atlantic-origin layer; and a cold, saline deep layer separated by transition zones. A strong inverse thermocline, where temperature increases rapidly with depth, is found between the upper and Atlantic layers, and a weak thermocline is found between the Atlantic and deep layers. Within this simple salt-stratified ocean, two distinct water mass assemblies are found—the Western Arctic assembly (WAA) and the Eastem Arctic assembly (EAA). These two

are distinguished, in the upper layer, by the presence or absence of Pacific-origin water (McLaughlin et al., 1996).

Water mass properties of the WAA in the Arctic Ocean are characterized by Pacific-origin water and, since 1993, are found only in the Canada Basin. The WAA upper layer contains polar mixed water to a depth of about 50 m, and a halocline layer which extends to about 200 m. Upper layer waters overlie the Atlantic layer whose warm core is found near 500 m. In contrast, EAA waters are characterized by an absence of Pacific-origin water in the upper layer and, since 1993, are found in Nansen, Amundsen, and Makarov basins. The EAA upper layer is comprised of a polar mixed layer that extends to 50 m, and a halocline layer which extends to 120 m. The EAA upper layer overlies the Atlantic layer with a warm core found near 275 m.

The Canada Basin halocline has three distinct components. Two of these components are dominated by Pacific-origin water which enters via Bering Strait and crosses the wide Chukchi Shelf where it is seasonally modified by productivity and ice formation. Inflowing Pacific-origin waters Include nutrient-rich water from the Gulf of Anadyr and fresher Alaskan Coastal Water characterized by low nutrients (Walsh et al., 1989). Seasonal modification on the Chukchi Shelf produces two forms of Pacific-origin water, referred to as Bering Sea summer and winter waters (Coachman and Bames, 1961). Summer inflow, augmented by riverine and ice- melt waters, is fresher and characterized by a local temperature maximum between S=31 and S=32, low nutrient concentrations and high oxygen concentrations. Winter inflow, augmented by ice formation, is more saline than summer inflow and characterized by a temperature minimum near S=33.1, high nutrient concentrations, and low oxygen concentrations. Dense water produced by ice formation on the Mackenzie Shelf, from water sufficiently preconditioned by storm- induced upwelling, also contributes to the maintenance of the temperature minimum (Melling and Moore, 1995).

The third component of the Canada Basin halocline is water of Atlantic- origin. It is formed either on the Barents Sea shelf (Jones and Anderson, 1986;

(Salmon and McRoy, 1994; McLaughlin et al., 1996; Guay and Falkner, 1997)). In the Canada Basin, this third component of the halocline is characterized by both low nutrient and low oxygen concentrations, resulting in a local minimum in NO (defined by Broecker (1974) as NO= 9NO3 + O2) at S=34.4 (Jones and Anderson, 1986). Although these waters have been labelled in the literature as upper, middle and lower halocline, these terms do not identify whether they are of Pacific (i.e. WAA) or Atlantic (i.e. EAA) origin.

Below the Canada Basin halocline lies the Atlantic layer, identified by a temperature maximum at its core. To reach the Canada Basin, Atlantic layer waters pass through three upstream basins, during which time this water becomes colder and fresher. Because Pacific-origin water adds to the thickness of the halocline in the WAA, Atlantic layer water lies deeper in the Canada Basin water column than in the upstream basins. Recent temperature and salinity measurements in the eastem Nansen Basin (Schauer et al., 1997) identified two distinct branches of Atlantic water in the Arctic Ocean; one entering through Fram Strait (FSB), and one seasonally modified during transit across the Barents Sea shelf and entering primarily through St. Anna Trough (BSB). These observations confirmed the two-branch hypothesis proposed by Anderson et al. (1994) and Rudels et al. (1994). Schauer et al. (1997) demonstrated that BSB inflow water, when compared to FSB water, was colder and fresher in the top 700 m and warmer and fresher below 800 m (their Figure 8). Downstream of the confluence zone, and north of the Laptev Sea, evidence of vigorous mixing between these two branches was manifested in temperature-salinity properties. Composition of the Atlantic layer, exiting eastward from the Amundsen Basin in 1993, was estimated by Schauer et al. (1997) to be 60% FSB near 275 m (EAA Atlantic layer core depth) and 80% BSB between 800 m and 1000 m.

The term “Atlantic layer” is used throughout this discussion to denote origin and to signify that this layer is comprised of two components that vary in relative amounts—the FSB, whose presence is identified by its Omax and found higher in the

water column, and the BSB, whose presence Is Identified by Its freshness and found lower In the water column. The Atlantic layer so-defined extends to about 2000 m In the Canada Basin and encompasses much of the weak thermocline. This definition expands the classical description of Coachman and Aagaard (1974) who characterized Atlantic Water as water warmer than 0.0 °C. Physical characteristics of FSB and BSB water, as well as their rates of Inflow Into the Arctic Ocean’s Atlantic layer, will vary from year to year according to seasonal and atmospheric conditions.

3.2 Temperature and salinity observations

Potential temperature and salinity CTD measurements, collected between 1989 and 1995 at Station A In the Canada Basin, were examined for evidence of change using the basic water mass layer approach described above (Figures 2a, 2b, and 2c). In the water column’s upper layer, the mixed layer was found to exhibit significant variability between years, without a temporal trend (Figure 2a). Surface salinity varied from less than S=25, when temperatures were 0.5 °C to -1.3 “C, to greater than S=28 when temperatures were near -1.5 °C. These measurements reflected annual freshwater variation from both local (Mackenzie River outflow) and upstream (Alaska Coastal Current, Bering Sea and Pacific Ocean) sources, as well as annual variations In atmospheric conditions affecting Ice cover and surface drift.

Below the mixed layer In the upper halocline, annual variability In temperature and salinity of the local 0max was observed without trend. Salinities ranged between S=31.35 and S=32.35 where the local temperature maximum was between -1.10“C and -1.40°C at depths from 55 m to 80 m (Figures 2b and 3). The range In physical properties, like observations made by Coachman and Bames (1961), showed that temperature and salinity In the upper halocline are seasonally conditioned. Winter Ice production and export, which vary annually. Influence the depth of the polar mixed layer and. In tum. Impact on temperature and salinity. Summer Import of Alaska Coastal Current and Bering Sea waters, whose temperatures and salinities vary from year to year, likewise influences characteristics of the upper halocline.

0.4

0.2

0.0

2

0.8

1.0

Salinity

Figure 2a. CTD potential temperature versus salinity from Canada

Basin Station A, 1989-1995: 8= 2 5 to S=35.

0.0

0.6

0.8

Salinity

Figure 2b. CTD potential tem perature versus salinity from Canada

Basin Station A, 1989-1995; 8 = 3 1 .0 to 8=34.7.

27.9

27.95

28.0

28.05

0.1

Salinity

Figure 2c. CTD potential temperature versus salinity from Canada

Basin Station A, 1989-1995: 8= 34.75 to 8=34.96

Next, in the middle layer of the halocline, the local temperature minimum was obsen/ed at salinities from S=33.0 to S=33.1. Temperatures were colder than -1 .5 0 °C in 1989 and 1990, and warmer than -1 .4 8 °C in 1992,1993 and 1995. Between 1989 and 1992, depth of the temperature minimum ranged from 200 m to 170 m. In 1993 and 1995, however, the temperature minimum was shallower and obsen/ed at 140 m.

Together with the mixed layer, the upper and middle haloclines constitute the sum of Pacific-origin water in the Canada Basin. A measurement of the contribution of Pacific-origin halocline water in the water column can be estimated simply by subtracting the depth of the 0min from the depth of the local 0oiax- Between 1989 and 1992, the depth between the local 0ma% and 0min ranged from 80 m to 140 m. By 1993 and 1995, however, the depth decreased to 65 m, signalling that a reduction in the amount of Pacific-origin water had occurred. This reduction could be explained by either reduced Pacific or enhanced Atlantic inflow, given that the boundary between Pacific-origin and Atlantic-origin waters in the Canada Basin lay below the middle halocline between S=33.1 and S=34.4. It could also be explained by displacement within the Canada Basin.

In the lower halocline at S=34.4, a distinct shift in temperature was also evident over time. Between 1989 and 1992, temperatures cooled slightly from -0.45 “C to -0.48 °C, however, a marked decrease of 0.14 °C occurred between 1992 and 1993. This temperature decrease was evident in the salinity range S=33.6 to S=34.4 in 1993. In 1995 the temperature remained unvarying at -0.62 °C. Temperature and salinity profiles (Figures 3 and 4) showed that this colder, denser lower halocline lay 40 m higher in the water column in 1993 and 1995.

Farther down the water column, a clear difference in temperature was also apparent between 1992 and 1993 in both components of the Atlantic layer (Figure 2c). In 1993, the upper FSB component, defined by 0m«, was slightly colder (from 0.43 °C to 0.40 °C), fresher (from 8=34.860 to 8=34.830), and found higher in the water column (from 560 m to 450 m). Temperature-salinity values

Thêta'"C

0.0

0.5

-0.5

B

200

300

400

Thêta

0.4

0.5

500

1500

2500

3000

3500

Figure 3. CTD potential temperature profile at Canada Basin

Station A, 1989-1995.

Salinity

25.0 26.0 27.0 28.0 29.0 30.0 31.0 32.0 33.0 34.0 35.0

0

a .

300

400

Salinity

34.80 34.82 34.84 34.86 34.88 34.90 34.92 34.94 34.9(,

500

1500

1995

2000

₁₉₉₃

2500

3000

3500

differed appreciably from prior years and were reflected also in tfie temperature profile (Figure 3). In 1993, below tfie 0max. temperature decreased sharply from 0.40°C to 0.37°C between 460 m and 550 m, was highly variable (temperature spikes) for the next 80 m, then decreased with depth at the same slope as other years. In 1995, 0max occurred at the same salinity, depth, and temperature as in 1993, however the temperature profile decreased smoothly below 0max from 450 m to 650 m.

The lower BSB component of the Atlantic layer, characterized primarily by lower salinity, was found to be slightly fresher between 1200 m and 2500 m in 1993 than in previous years (e.g. at 1600 m AS=-0.004 from 8=34.929 to 8=34.925). Freshening was paralleled by a slight increase in temperature (e.g. at 1600 m A0=O.O1O) between 1400 m and 2000 m. These small variations were not apparent in the 1993 temperature-salinity plot in the range 8=34.890 and 8=34.950. By 1995, however, the B8B component of the Atlantic layer appeared markedly different in the temperature-salinity plot. Figure 2c illustrates that freshening, evident only in the F8B component in 1993 between 8=34.830 and 8=34.875, extended in 1995 from 8=34.830 to 8=34.910 and included both F8B and B8B components of the Atlantic layer. When viewed by depth, salinity was markedly fresher between 500 m and 2500 m in 1995 (e.g. at 1000 m a8=-0.011). Likewise,

the 1995 temperature was warmer between 700 m and 2200 m (e. g. at 1000 m A0=O.O46 °C). Temperature and salinity profiles show that, below the F8B 0ma% between 500 m and 2500 m, cold and salty water was replaced by warmer and fresher water in 1995. Further, in 1995, the B8B component was evident to 2500 m and water of salinity 8=34.880 was found about 150 m deeper. This suggested that the B8B component of the Atlantic layer occupied more of the water column than in previous years.

In summary, analysis of temperature and salinity measurements demonstrated that distinct changes were first manifest in 1993 over much of the Canada Basin water column at Station A. Differences observed were: shallowing of the middle halocline and thinning of Pacific-origin water; shallowing and cooling of the lower halocline; shallowing, cooling and freshening of the F8B; and warming.

freshening and an Increased volume of the BSB. Beginning in 1993, the most significant modification was found in the Atlantic layer's BSB component. The 1995 temperature and salinity measurements between 500 m and 2700 m in the Canada Basin were similar to 1993 measurements between 500 m to 1500 m in upstream Eurasian Basin waters, reported by Schauer et al. (1997). This similarity pointed to the value of comparing upstream and downstream data sets. Physical and geochemical data from upstream stations will now be compared with Canada Basin data to determine if differences observed in the Canada Basin halocline and Atlantic layers could be traced to upstream BSB outflow.

3.3 Barents Sea outflow characteristics

BSB outflow properties were examined first by comparing physical measurements collected in the Nansen, Amundsen and Makarov basins in 1993 with measurements made in the Canada Basin from 1989 to 1995 (see Figure 1). Data from the western and eastern Nansen, and eastem Amundsen basin slopes at stations 21, 70, and 35 collected during the 1993 Polarstem ARK IVX/4 expedition (Schauer et al., 1997), from the southern Makarov Basin slope at stations E and 20 collected, respectively, during the 1993 CCGS Larsen expedition (McLaughlin et al., 1996) and the 1994 CCGS Louis S. St. Laurent Arctic Ocean section (Swift et al., 1998), were combined in a temperature-salinity plot (Figure 5). Temperature and salinity measurements from one Nansen Basin station upstream (Station 21 ) and one station downstream (Station 70) of the St. Anna Trough, presented by Schauer et al. (1997), identified the physical characteristics of the two distinct branches of Atlantic water— FSB and BSB— in the Nansen Basin. As Figure 5 shows, both FSB and BSB waters extended over a range of temperature and salinity values, signalling the presence of a water mass, not a water type (i.e. water defined by a single temperature or salinity). Measurements from downstream stations in the Amundsen (Station 35) and Makarov basins (Stations E and 20) indicated that water from both branches mixed with each other and with older, ambient basin waters, and that the Atlantic layer was comprised of two components, each reflecting attributes unique to their source waters. The FSB

/ N B 21

FSB

AB 3 5

MB 20

2

0.8

28.05

CB '89-95

7

-.78 34.80 34.82 34.84 34.86 34.88 34.90 34.92 34.94 34.96 34.98

Salinity

Figure 5. CTD potential temperature versus salinity:S=34.78 to 8 = 3 4 . 9 8 .

Data are from Ca nad a Basin Station A (1 989-19 95); Nansen

Basin Stations 21 and 70, (1993); Amundsen Basin Station 35

component, characterized by 9max. began warm and saline and became progressively cooler and fresher along an isopycnal surface (i.e. oe =27.9) as it travelled from Nansen to Amundsen and Makarov sub-basins. Conversely, the lower BSB component, characterized by colder and fresher water, became warmer and more saline following an isopycnal surface (i.e. oe =28.0) as it travelled from the Nansen to the Amundsen Basin and, thereafter, decreased only slightly in temperature and salinity from the Amundsen to the Makarov Basin.

Temperatures and salinities obsen/ed in the Canada Basin's Atlantic layer presented a smooth, round contour significantly different in shape than temperature-salinity contours observed in upstream basins. The shape of the temperature-salinity plot suggested that the extent of mixing between the Atlantic layer's two components with ambient water in the Canada Basin was far greater than found in upstream basins, where the average characteristics of inflowing waters over time were reflected. Notwithstanding this mixing, FSB and BSB components of the Atlantic layer remained identifiable in the Canada Basin by a 8max and an inflection in salinity near S= 34.890 respectively. The upstream basin- to-basin progression of FSB's 0max along the isopycnal surface oe =27.9 did not continue into the Canada Basin where e^ax was observed at oe >27.95. The fact that 0max was found at a higher density in the Canada Basin than it was upstream denoted long-term temporal variability in the characteristics of Fram Strait inflow (Loeng ,1990) and the recent warming of Atlantic inflow.

Beginning in 1993, temperature and salinity characteristics in the Canada Basin's Atlantic layer appeared to be influenced by changes in Barents Sea outflow (Figure 5, Station 70), because the FSB component became fresher and colder, and the BSB component became fresher and occupied more of the water column. These observations suggested that composition of the Atlantic layer in the Canada Basin had been altered to include a higher volume of colder, fresher BSB water, evident from 400 m to 2500 m of the water column. Upstream and downstream temperature and salinity measurements, viewed over a larger salinity scale (Figure 6), showed that modification to both halocline and Atlantic layers in the Canada Basin water column was strongly associated with an increased Atlantic-origin

2.4

2.2

2.0

NB 21

NB 70 //:

BSB /

AB 35

0.4

MB E

0.2

0.4

CB 89-95

33.8 33.9 34.0 34.1 34.2 34.3 34.4 34.5 34.6 34.7 34.8 34.9 35.0

Salinity

Figure 6. CTD potential temperature versus salinity: 8 = 3 3 . 8 to 8 = 3 5 . 0 .

Data are from Canad a Basin 8tation A (1989- 19 95 ); Nansen

Basin 8tations 21 and 70 (1993); Amundsen Basin 8tation 35

(1993); and Makarov Basin 8tations E (1993) and 20 (1994).

Atlantic layer changes in the Canada Basin, geochemical measurements from upstream and downstream stations will now be compared.

3.4 Time-series geochemical observations

Geochemical measurements, collected at Station A in the Canada Basin from 1989 to 1995, were compared over time and compared with measurements from four upstream stations. Nutrient and ’^1 isotope values were analyzed primarily to investigate whether geochemical differences were evident in the halocline and whether the source of these differences could be ascertained. Oxygen and CFC measurements were then analyzed to answer questions about the Atlantic layer’s composition and to distinguish between non-shelf (FSB) and shelf-derived (BSB) water.

Nutrients and ’^1 were selected to examine the boundary separating Atlantic and Pacific-origin waters within the halocline. Nutrients define the Atlantic/Pacific water mass boundary because Pacific-origin water is high in nutrients, and Atlantic- origin water low. Conversely, ’®l defines the Atlantic/Pacific boundary equally effectively because it signals the presence of Atlantic-origin water. Since the 1960s, Atlantic water has been enriched by '^1 radionuclide emission from European reprocessing plants, whereas Pacific-origin water, in contrast, exhibits low fallout ’* ! concentrations (Smith et al., 1998). Although Canada Basin time-series ’^1 measurements dating back to 1989 are not available, ^^1 measurements from Canada and Makarov basins in 1993 and 1995 were used in conjunction with time- series nutrient measurements beginning in 1989 to determine the depth of the Atlantic/Pacific water mass boundary over this six-year period.

Analysis of source waters in the middle and lower halocline of the Canada Basin at Station A began by plotting silicate, '^1, and salinity profiles from 1995 measurements (Figure 7). The silicate maximum, found near S=33.1, confirmed the presence of Pacific-origin water in the middle halocline. Likewise, the

q

Silicate mmol/m

0

5

10

15

20

25

30

35

40

1 I I I r I I I I I I I r r r I r I I I I | | | | | | | | i I i r i i I i i i i I

lodine-129 x1 o"* ^atoms/m^

Q . (D Q

20

25

30

10

15

35

0

5

40

0

100

300

400

500

600

29

30

31

32

33

Salinity_____

34

35

Silicate mmol/m

I

Silicate mmol/m

maximum observed near S=34.8, confirmed the presence of Atlantic-origin water in the lower halocline. Intersection of these two profiles pointed to the location of the Atlantic/Pacific boundary, which in 1995 occurred at 240 m and S=34.4 in the lower halocline. Time-series silicate profiles from 1989 to 1995 (Figure 8a) indicated three kinds of changes. First, the zone of Pacific-origin water, as outlined by the silicate maximum, decreased in thickness from 175-220 m, between 1989 and 1992, to 150 m in 1993 and 1995. Second, the silicate maximum was located about 50 m higher in the water column in both 1993 and 1995 than in previous years. Third, in 1995 the concentration of the silicate maximum was lower than in previous years. Altogether, these findings demonstrated that the boundary between Pacific and Atlantic-origin waters in the Canada Basin halocline was shallower in 1993 and 1995 than in earlier years and suggested a recent displacement of Pacific-origin water from below.

A shift in the Atlantic/Pacific boundary was also evident when Station A time- series silicate measurements were plotted against salinity (Figure 9a). Between 1989 and 1993, concentration of the silicate maximum near S=33.1 in the middle halocline ranged 36-38 mmol m"^ but, in 1995, decreased to 33.5 mmol m"^. Silicate decrease suggested either that more Atlantic-origin water was present due to shallowing of the Atlantic/Pacific water mass boundary, or that the silicate concentration of Pacific-origin BSW W had been reduced. Comparison of silicate measurements from the Nansen, Amundsen and Makarov basin stations in the same salinity range pointed to an increased contribution of Atlantic-origin water in Canada Basin water column below S=33.1. Atlantic-origin concentrations were much lower (5 mmol m"^) than concentrations in Pacific water (>33 mmol m"^). No data were available, however, to determine whether BSW W composition had been altered during this time.

Below the lower halocline, from S=34.4 to S=34.88, silicate concentrations in 1993 were also lower than previously observed. By 1995, lower silicate concentrations were observed from S=33.1 to S=34.88 in a region of the water column including the middle halocline, the Atlantic/Pacific boundary, the lower halocline, and both components of the Atlantic layer. Tfie 1995 obsenration of lower

C CD CO CO n (U CD < CD Ô)

c

o

D>

t>

Atlantic water now occupied more of the halocline than in previous years, and that the boundary separating Pacific and Atlantic-origin waters was located higher in the water column. Decrease in concentration of the silicate maximum in 1995 near S=33.1 identified the extent of Atlantic-origin water’s influence in the middle halocline.

Moreover, lower silicate concentrations in the water column at salinities from 8=34.40 to 8=34.88 in 1993 and 1995 suggested that differences in silicate concentration could also signal a modification to the composition of Atlantic layer water. At the two Nansen Basin stations, silicate concentrations were lower in 8 8 8 source water than in FS8 source water below 300 m (Figure 8b) and when salinities were greater than 8=34.80 (Figure 9b). Downstream in the Amundsen and Makarov basins, where both F 8 8 and 8 8 8 waters comprise the Atlantic layer, low silicate concentrations below the salinity maximum indicated the presence of 8 8 8 water. Farther downstream, in the Canada Basin, lower silicate concentrations were obsen/ed in 1995 to a depth of about 1200 m (8=34.89), and these lower concentrations suggested that composition of the Atlantic layer now included a greater amount of 8 8 8 water.

Analysis of nitrate (Figures 10, 11) and phosphate (Figures 12, 13) time- series measurements corroborated silicate findings from 8tation A, notably: the nutrient maximum of the middle halocline was shallower and thinner in 1993 and 1995 than in previous years; nitrate and phosphate concentrations were lower in 1993 at salinities 8>34.4 and, in 1995, lower at salinities 8>33.1; and in 1995 the maximum nitrate and phosphate concentrations were lower than in previous years. Moreover, nitrate and phosphate concentrations were also lower in 8 8 8 than in F 8 8 source water. These data signalled that the Atlantic/Pacific boundary was shallower and, accordingly, more Atlantic-origin water was found within the Canada Basin halocline. Lower concentrations of nitrate and phosphate at salinities 8>34.4 pointed to an increase of 8 8 8 water in the composition of Atlantic layer.

0

400

600

800

£

§ 0 0 0

1400

1600

1800

2000

0

2

4

Nitrate mmol/m^

6

8

10

12

14

16

A

A

A

A

A

♦

♦

♦

♦

♦

♦

♦

♦

♦

♦

18

CO

18

17

16

15

14

13

12

E 11

E

E

A

A

A

A

A

i

È

31.0

31.5

32.0

32.5

33.0

33.5

34.0

34.5

35.0

S a lin ity

0

200

400

600

800

E

§ 0 0 0

1400

1600

1800

Phosphate mmol/m

0.5

1.0

1.5

^ — I— I— I— I— I— I— I

A A ^ a a A A ^ a

^

Z

A

A

A

A

A

CO

2.1

-1.9 -

-1.7 -

1 . 6

1.5

-J 1.4

H

E

E

1.3

I.O

O 0.9 H

0.7 -

0.5 —

0.4

0.3

H

0.2

A

A

A

A

A

È

31.0

31.5

32.0 32.5

33.0

33.5

34.0

34.5

35.0

__________________________S a lin ity _________________________ ,

Canada Basin time-series oxygen profiles (Figure 14) at Station A also provided corroborating evidence that more Atlantic-origin water was found in the Canada Basin halocline. In 1993 and 1995, the oxygen minimum of the lower halocline at 210 m was 50 m shallower, and the concentration was 15 mmol m'^ higher than in previous years. By 1995, oxygen concentrations were higher at depths from 200 m to 1000 m, suggesting that these waters were more ventilated than in the past. The oxygen-salinity plot (Figure 15) showed that oxygen concentrations were higher between salinities S=33.1 and 8=34.9. The plot also illustrated that oxygen concentrations throughout the water column in the three upstream basins were much higher than in the Canada Basin and, in addition, that BSB shelf source water was about 12 mmol m"^ higher than FSB source water. Thus, higher oxygen concentrations within the Canada Basin water column also pointed to the increased contribution of shelf-source (BSB) water within the Atlantic layer.

But oxygen is not a consen/ative tracer and cannot unequivocally identify shelf-source water. The conservative tracer CFC-11, measured at Station A in 1992,1993 and 1995, however, provided more definitive identification. Since CFC-11 concentrations denote levels of ventilation, this tracer was used to identify shelf-source water, which exhibits high levels of ventilation at depth within the water column as a result of recent convection, from non-shelf source water. Furthermore, because CFC-11 solubility increases as water temperature decreases, this tracer's value in distinguishing cold, BSB shelf-water from warm, non-shelf FSB water is enhanced. CFC concentrations in the water column thus reflect both the atmospheric concentration at time of subduction (i.e. age) and the water’s local temperature (i.e. source).

Comparison of CFC-11 measurements at Station A from 1992, 1993, and 1995 showed that the water column from 200 m to 1600 m was markedly more ventilated in 1995 and that the largest increase in CFC-11 concentration occurred in the lower halocline and the Atlantic layer between 300 m and 1000 m (Figure 16). Similarly, a CFC-11-salinity plot (Figure 17) demonstrated that the largest

240

0

400

600

800

1400

1600

1800

2000

Oxygen mmol/m

300

320

340

360

380

♦

O

♦

&

♦

♦

♦

A

A

A

A

A

400 n

380

-360

340

E

320

Q 300

280

260

-240

A

A

A

A

A

♦

r—I—i—i—I I I I I I—I—I—I—I—I—I—I I I I 1—r—1—I—I—I—i—i—I—I—I—I—I—I—I—I—I—[—1—I

31.0

31.5

32.0

32.5

33.0

33.5

34.0

34.5

35.0

________________________________ S alin ity_________________________ ,

CFC-11 nmol

0

1

2

3

4

5

6

500

1500

?

♦

#

A

Figure 16. CFC-11 profiles

» 6 >

5

E

%

3

-0 ---

34.0 34.1 34.2 34.3 34.4 34.5 34.6 34.7 34.8 34.9 35.0

Salinity